Abstract
Keywords
Introduction
Open surgery remains the gold standard in aortic arch repair for aneurysms or dissections. Various strategies have been developed to limit the morbidity and mortality associated with open surgical repair, the major concern being the reported 5% to 12% incidence of perioperative stroke.1,2 Selective cerebral perfusion associated with deep hypothermia during circulatory arrest appears to reduce neurological morbidity. 3 The frozen elephant trunk technique has been recently introduced to facilitate 1-stage repair of complex pathologies of the aortic arch and proximal descending aorta. 4 Despite these recent advances, aortic arch reconstruction remains challenging, particularly in elderly and/or polymorbid patients or those requiring emergency repair. Many patients are deemed unsuitable for open repair. 5
More recently, a combined endovascular and open approach has been adopted as a valuable alternative, consisting of supra-aortic debranching and revascularization followed by stent-graft deployment. 6 Debranching provides an appropriate landing zone for the stent-graft while preserving perfusion to the supra-aortic branches. Although this approach provides an attractive alternative for the treatment of aortic arch pathology, the majority of these adjunctive procedures are major operations that can have significant perioperative mortality. Branched stent-grafts that permit complete endovascular aortic arch repair have been proposed,7,8 but this approach has several limitations. Notably, the time required to manufacture custom-made stent-grafts precludes their use for urgent cases, and their high costs also have to be considered. Most important, there is a high rate of embolization associated with this approach, probably related to the complexity of deploying a multibranched unibody stent-graft. 8
An alternative option is a physician-modified thoracic stent-graft with a proximal scallop or single fenestration for treating distal arch pathology where the proximal landing zone commences at the distal margin of the left common carotid artery (LCCA) and blood flow into the left subclavian artery (LSA) needs to be preserved. Furthermore, such modified thoracic stent-grafts can also be used to preserve flow into the LCCA or the innominate artery if the more distal arch branches are revascularized by other means, such as bypass surgery. Some centers have reported good short-term results for juxta- and suprarenal aneurysms,9,10 but there are no data on their applicability for aortic arch lesions. Moreover, there are concerns regarding feasibility and durability.
The aim of this study was to assess outcomes of physician-modified (single fenestration or proximal scallop) thoracic stent-grafts for the treatment of aortic arch lesions in high-risk patients.
Methods
Patient Cohort
Between November 2013 and June 2016, 95 patients with aortic arch lesions were treated in 2 tertiary referral centers (A de Villeneuve Hospital, Montpellier, France; Hakodate Municipal Hospital, Hakodate, Japan). During this period, 35 patients underwent open aortic arch repair, and 24 patients had hybrid arch repair. Thirty-six patients (mean age 74.7±9 years, range 58–91; 27 men) were treated using physician-modified thoracic stent-grafts [34 Valiant (Medtronic Vascular, Santa Rosa, CA, USA) and 2 Zenith TX2 (Cook Inc, Bloomington, IN, USA)] on a compassionate-use basis. Patient demographics are listed in Table 1. Half of the patients had a degenerative aneurysm (1 ruptured and 1 symptomatic). The remainder had type B dissection (n=9), traumatic transection (n=3), type Ia endoleak after previous thoracic endovascular aortic repair (TEVAR) (n=5), or aortoesophageal fistula (n=1). All patients were considered to be at high surgical risk owing to serious comorbidities (American Society of Anesthesiologists score ≥3) or emergent repair (11, 30%). The protocols and informed consent documents for treatment were approved by the respective institutional review boards.
Patient Characteristics. a
Abbreviations: COPD, chronic obstructive pulmonary disease; TEVAR, thoracic endovascular aortic repair.
Continuous data are presented as the means ± standard deviation; categorical data are given as the counts (percentage).
Patient Eligibility
Our policy is strictly the same for physician-modified or standard thoracic stent-grafts (a minimum aneurysm diameter of 5.5 mm is required). Patients with aortic arch lesions were eligible for TEVAR using a physician-modified thoracic stent-graft if the proximal and distal aortic neck diameters were <40 mm, the proximal neck length distal to the LSA was <15 mm, and the distal neck length was >15 mm and healthy (nondissected).
Patients were suitable for a scalloped stent-graft only if the minimum healthy aortic seal length was at least 15 mm from the expected position of the edge of the scallop, both proximally and laterally. If the healthy aortic seal length was <15 mm lateral to the target vessel, patients were candidates for TEVAR using a fenestrated stent-graft. A <5-mm distance between the innominate artery and the LCCA precluded the use of a fenestration.
Planning, Sizing, and Device Preparation
All patients underwent high-resolution computed tomography angiography (CTA) preoperatively. Procedure planning and device sizing were performed using a dedicated 3-dimensional (3D) vascular imaging workstation [Aquarius (TeraRecon, Foster City, CA, USA) or OsiriX Imaging Software package (Pixmeo SARL, Bernex, Switzerland)]. Centerline lumen reconstructions were used to determine aortic diameter at the proximal and distal landing zones and the distance between the proximal edge of the stent-graft and the center of the fenestration or scallop. Axial imaging was used to assign clock positions to the origins of each branch vessel using centerline reconstructions with the innominate artery at the 12 o’clock position. Stent-grafts of sufficient length were selected to provide minimum 15-mm proximal and distal landing zones in healthy aorta. Stent-graft oversizing was <10% for acute aortic dissection and between 10% and 15% for other aortic arch pathologies. Volume-rendering images were used to determine the optimal position of the C-arm and to evaluate aortic arch tortuosity.
Stent-Graft Modification
The stent-grafts were individually designed to fit the aortic configuration of each patient. Modification of the stent-graft was performed on a back table, commencing before the start of anesthesia. A portion of the device was unsheathed, preferably the area to be modified plus an additional stent. The fenestrations and scallops were premarked on the main stent-graft according to the measurements obtained from centerline analysis.
Proximally Scalloped Stent-Graft
The length of the scallop matched the distance between the adjacent supra-aortic branch and the distal edge of the target vessel ostium. The width of the scallop was that of the target vessel plus 5 mm on both sides. Construction of the scallops has been reported. 11
Fenestrated Stent-Graft
A single fenestration for the supra-aortic target vessel was made between the stent struts. Fenestrations were circular, without stent struts going across them, and 5 mm larger than the target vessel. The location of the fenestration was marked on the stent-graft based on both the length and clock face measurements determined using the reconstructed images; minor adjustments were permitted to avoid struts. A larger fenestration was made when the lesion was in the inner curvature of the aortic arch to avoid rotation problems during delivery. A cautery device was used to carefully burn the Dacron fabric; a radiopaque nitinol wire was sewn onto the edge of the fenestration for reinforcement.
Technique
All procedures were performed by team surgeons in an operating room equipped with a C-arm or in a hybrid room with the patient under general anesthesia. If planned, supra-aortic trunk revascularization preceded TEVAR in a single stage procedure.
A surgical cutdown was made over the common femoral artery, and heparin (5000 units) was administered as the thoracic stent-graft was introduced over an ultra-stiff guidewire. Angiography was performed via a pigtail catheter introduced percutaneously through the contralateral common femoral artery. Mean blood pressure at deployment was lowered to ~80 mm Hg to optimize accuracy. The delivery system was introduced so as to orient the markers on the stent-graft fenestration or scallop to the superior aspect of the arch. If not properly oriented, the stent-graft was pulled back in the descending thoracic aorta and rotated to adjust the position. Once properly positioned, the stent-graft was partially unsheathed. Using the optimal C-arm position determined preoperatively on the 3D reconstruction, an angiogram “opening” the aortic arch was used to precisely align the radiopaque marker with the target vessel. Once the fenestration or the scallop was oriented toward the supra-aortic target vessel, the stent-graft was fully deployed. Minor rotational adjustments were possible to fine tune the position of the fenestration once the first stents were deployed. Bridging covered stents (iCAST; Atrium Maquet Getinge Group, Hudson, NH, USA) to target vessels were used selectively in cases of imperfect alignment of the fenestration or in case of nonapposition of the fenestration to the aortic wall.
Follow-up surveillance was performed with serial CT scans at 1 week, then at 3, 6, and 12 months, and annually thereafter. A duplex scan was performed in case of clinical or CT abnormality.
Results
Twenty-four (67%) patients were treated using an aortic arch stent-graft with a single homemade fenestration (Figure 1), while 12 (33%) patients received a modified proximal scalloped stent-graft (Figure 2). Twenty-four (67%) patients had supra-aortic trunk revascularization (LCCA and LSA in 14 and LSA in 10) via a cervical approach prior to stent-graft delivery. Bridging covered stents were used in 5 cases. The mean proximal sealing length was 23.3±5 mm. Zone 0 was involved in 16 patients, zone 1 in 9, and zone 2 in 11 (Table 2).
(A) The Valiant thoracic stent-graft was partially unsheathed, and the proximal fenestration was fashioned on the upper side of the aortic arch. (B) An 84-year-old patient with aortic arch aneurysm underwent a crossover bypass from the right axillary artery to left carotid artery and to the left axillary artery. (C) Completion angiography after implantation of the proximally fenestrated stent-graft in zone 0 demonstrates exclusion of the arch aneurysm and patency of the innominate artery.
(A) The Valiant thoracic stent-graft was partially unsheathed, and a proximal scallop (arrow) was fashioned on the upper side of the aortic arch. (B) A 78-year-old patient with aortic arch aneurysm after acute type A aortic dissection repair underwent a crossover bypass from the right common carotid artery to left subclavian artery and reimplantation of the left common carotid artery into the bypass. (C) Completion angiography after implantation of the proximally scalloped stent-graft in zone 0 demonstrates exclusion of the arch aneurysm and patency of the innominate artery.
Procedure Details. a
Continuous data are presented as the means ± standard deviation; categorical data are given as the counts (percentage).
Perioperative Outcomes
Mean time required for stent-graft modifications was 18 minutes (range 14–21). Technical success with sealing was obtained in all cases and no type I endoleak. All the arteries were patent, though 1 patient had a 50% stenosis of the LSA secondary to positioning the stent-graft too laterally. The patient is asymptomatic and being followed closely.
One (3%) patient in the landing zone 1 group had a stroke, probably due to atheroembolism during LSA transposition or stent-graft placement; there were no permanent sequelae. One patient developed a retroperitoneal hematoma due to sheath introduction during stent-graft placement. Two patients had dehiscence of their groin wounds requiring surgery. One patient had an infected carotid-carotid bypass graft requiring graft explantation and in situ revascularization using a great saphenous vein.
Two (6%) patients died within 30 days. One experienced a rupture of the descending thoracic aorta (confirmed by autopsy) on the first postoperative day; there was no type I endoleak seen at completion angiography. Another patient died on the fifth postoperative day from the sequelae of traumatic rupture of the left kidney (grade 4 lesion), which was managed conservatively and unfortunately had delayed massive hemorrhage. Mortality related to aortic repair was 3%. No paraplegia or retrograde dissection was recorded in the perioperative period.
Follow-up
During a mean follow-up of 11.4±6 months (range 2–36), there were no conversions to open surgical repair. Three additional patients died (14% overall mortality); one case was related to the aorta (6% overall aorta-related mortality). One patient treated for a ruptured arch aneurysm (zone 0 repair; innominate artery single fenestration) died of a rupture of the aortic arch 5 months after the procedure due to a type Ia endoleak related to aneurysmal evolution of the proximal neck. Three patients underwent successful embolization for type II endoleaks. One patient had an asymptomatic occlusion of the LSA and LCCA revascularization. No paraplegia or retrograde dissection was recorded during follow-up.
Discussion
This dual-center experience demonstrated the feasibility of physician-modified thoracic stent-grafts for the treatment of aortic arch lesions over a wide spectrum of thoracic aortic pathology. This approach provides a rapid method of modifying the stent-graft, which was particularly beneficial in the 11 emergent cases. The 6% aorta-related mortality, 97% supra-aortic branch patency (bypasses and fenestrations/scallops), and no early type I endoleak were encouraging short-term results. Only 1 late proximal endoleak was observed, which led to aortic rupture 5 months after TEVAR.
Although intentional endograft coverage of the LSA was initially thought to be a viable alternative in this setting, expanding experience has shown a significantly increased risk of subclavian steal syndrome, arm claudication, vertebral territory stroke, and spinal cord ischemia. 12 Thus, maintaining patency of the LSA will improve outcomes of emergent TEVAR. Insofar as zone 1 is concerned, a review of clinical outcomes in hybrid procedures for aortic arch lesions pointed out that zone 1 debranching was associated with 7.6% mortality, lower than zone 0 (15.1%, p=0.021) but still substantial. 13 Zone 0, requiring sternotomy and partial cross-clamping of the ascending aorta, is an invasive procedure, which is why a less invasive approach is being investigated for this subgroup of patients.
Major advantages of aortic arch stent-grafting with a single fenestration or scallop compared to branched stent-grafts are the simplicity and rapidity of the procedure, fewer manipulations in the arch, and a lower risk of microembolic strokes. 7 Only 1 (3%) patient in our cohort had a stroke without permanent sequelae.
There are special anatomical considerations when utilizing this technique. An aneurysmal ascending aorta is one of the most important factors precluding endovascular arch repair. Also, <5 mm between the innominate and the LCCA prevents the use of a zone 1 fenestration. Ascertaining that a fenestration or scallop is well oriented toward the supra-aortic target vessel is crucial.
Because of aortic arch geometry, a much longer length of aorta needs to be covered in the outer curvature to achieve an appropriate sealing zone, which can cause a bird-beak configuration of the proximal stent-graft at the inner curve. This is the etiology for the majority of TEVAR failures at this level, with most type Ia endoleaks originating from the inner curvature of the aortic arch. The concept of a proximal scallop for TEVAR allows simple extension of the proximal sealing zone (minimum 15 mm of healthy aorta both proximally and laterally) along the inner curve, while leaving supra-aortic trunks patent on the outer curvature.
If the healthy aortic seal length was <15 mm lateral to the target vessel, patients were candidates for TEVAR using a fenestrated stent-graft. Fenestrated TEVAR is based on the principle of implanting a stent-graft with fenestrations that will align perfectly to the target vessels to extend the proximal landing zone. The diameter and morphology of the aorta at the relevant segment must also allow the fenestration to be close to the target vessel. In patients with aneurysms involving the greater curve of the arch or when the stent-graft cannot appose the aortic wall at the level of the target vessels, a branched endograft is more suitable to achieve seal. This is the main technical limitation of the use of physician-modified thoracic stent-grafts. However, modified branched stent-grafts for thoracoabdominal aortic aneurysm are currently being investigated and could be proposed for the aortic arch. 14
All of these systems are customized, requiring accurate preoperative planning. Precise deployment of these fenestrated arch stent-grafts is important to correctly orient the fenestrations toward the target branches. Because of the homemade nature of these modifications, graft rotation and misalignment of the fenestration/vessel ostium interface can still occur. In addition, the aorta may change configuration after insertion of the semirigid stent-graft and thus alter alignment of the side branches. However, this was not observed in our experience.
Even if performed in sterile conditions, deploying the stent-graft on a back table for modification could be associated with an increased risk of infection. This complication was not observed in the present series. Other disadvantages include the necessity for the physician to spend time modifying the stent-graft and lack of industrial quality control after device modification. Modification of commercially available devices by physicians voids any guarantee of safety by the manufacturer.
Custom-made inner branched stent-grafts are currently available. Haulon et al 8 reported the world experience in a multicenter study involving 38 branched arch devices; they observed a 13% mortality, a 16% stroke rate, and a 15.8% technical failure rate, with secondary procedures necessary in 19.6%. Factors such as the delay in device manufacturing, anatomical and technical limitations, and expense limit the widespread uptake of this technology, which is additionally unsuitable for emergent cases. Above all, the technical difficulty of side branch catheterization results in an inherently high risk of cerebral embolism.
Centers in Japan have reported extensive experience with the use of custom-made arch endografts having large fenestrations that increase the margin of safety during deployment. In a recent study, 15 this technique was used in patients with a short seal zone distal to the LCCA (mean 11 mm, range 5–15); this resulted in a 32.4% type Ia endoleak rate at discharge and 16.2% aneurysm enlargement at follow-up.
The use of readily available “off-the-shelf” endovascular materials would decrease the cost of endovascular repair and allow the use of these minimally invasive techniques in a greater number of patients. Off-the-shelf branch devices (eg, Valiant Mona LSA; Medtronic Vascular) 16 consist of a main stent-graft and a branch stent-graft designed to maintain patency of 1 supra-aortic trunk while excluding the encroaching aneurysm, which could prove useful in emergent TEVAR cases. However, these devices are not currently on the market.
The chimney graft is yet another competing endovascular technique that preserves flow into endograft-covered aortic arch branches. In a recent review 17 of 182 patients who underwent chimney graft deployment during endovascular arch repair, primary technical success was achieved in 98%, with a 5.3% stroke rate and a 13.6% type Ia endoleak rate. Questions remain regarding durability and material interaction in the long term.
We acknowledge that the current practice of stent-graft modification is controversial. The fragility of homemade stent-grafts is a crucial problem, and the question of fabric durability still needs to be evaluated. Metal fatigue and material deterioration are known complications of endografting. These alterations, especially the use of a proximal scallop, might have an impact on the general stability of the graft. In our series, no stent fractures were detected by routine radiological examinations in follow-up. Careful long-term monitoring of patients is required to avoid major complications resulting from limited durability of these devices.
Until an “off-the-shelf” device is available, patients with rapidly expanding, symptomatic, or ruptured arch aneurysms who are poor candidates for open surgical repair have limited options other than immediate physician modification of an existing stent-graft. Strict surveillance of these stent-grafts and modifications will be necessary in the long term to monitor and ensure durability of repair because of the potential for stent collapse or stent strut breakage.
Conclusion
The use of physician modified thoracic stent-grafts for the treatment of aortic arch aneurysms is both feasible and effective for supra-aortic trunk revascularization during TEVAR for a spectrum of aortic arch pathology. Durability concerns will need to be assessed in additional studies with long-term follow-up.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.